Update README.md
Browse files
README.md
CHANGED
@@ -16,33 +16,74 @@ inference: true
|
|
16 |
|
17 |
This pipeline was finetuned from **runwayml/stable-diffusion-v1-5** on the **iamkaikai/amazing_logos_v4** dataset. Below are some example images generated with the finetuned pipeline using the following prompts: ['Simple elegant logo for Grupo Altair Publicidad, Circle Lines Venezuela, Publishing, successful vibe, minimalist, thought-provoking, abstract, recognizable, relatable, sharp, vector art, even edges']:
|
18 |
|
19 |
-
![val_imgs_grid](./val_imgs_grid.png)
|
20 |
-
|
21 |
-
|
22 |
-
## Pipeline usage
|
23 |
-
|
24 |
-
You can use the pipeline like so:
|
25 |
-
|
26 |
-
```python
|
27 |
-
from diffusers import DiffusionPipeline
|
28 |
-
import torch
|
29 |
-
|
30 |
-
pipeline = DiffusionPipeline.from_pretrained("iamkaikai/amazing-logos-v4", torch_dtype=torch.float16)
|
31 |
-
prompt = "Simple elegant logo for Grupo Altair Publicidad, Circle Lines Venezuela, Publishing, successful vibe, minimalist, thought-provoking, abstract, recognizable, relatable, sharp, vector art, even edges"
|
32 |
-
image = pipeline(prompt).images[0]
|
33 |
-
image.save("my_image.png")
|
34 |
-
```
|
35 |
|
36 |
## Training info
|
37 |
|
38 |
These are the key hyperparameters used during training:
|
39 |
|
40 |
-
*
|
41 |
-
*
|
|
|
42 |
* Batch size: 1
|
43 |
* Gradient accumulation steps: 1
|
44 |
* Image resolution: 512
|
45 |
* Mixed-precision: fp16
|
46 |
|
47 |
|
48 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
16 |
|
17 |
This pipeline was finetuned from **runwayml/stable-diffusion-v1-5** on the **iamkaikai/amazing_logos_v4** dataset. Below are some example images generated with the finetuned pipeline using the following prompts: ['Simple elegant logo for Grupo Altair Publicidad, Circle Lines Venezuela, Publishing, successful vibe, minimalist, thought-provoking, abstract, recognizable, relatable, sharp, vector art, even edges']:
|
18 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
19 |
|
20 |
## Training info
|
21 |
|
22 |
These are the key hyperparameters used during training:
|
23 |
|
24 |
+
* Dataset size: 400k
|
25 |
+
* Epochs: 2
|
26 |
+
* Learning rate: 1e-07
|
27 |
* Batch size: 1
|
28 |
* Gradient accumulation steps: 1
|
29 |
* Image resolution: 512
|
30 |
* Mixed-precision: fp16
|
31 |
|
32 |
|
33 |
+
|
34 |
+
![preview](./grid.jpeg)
|
35 |
+
|
36 |
+
|
37 |
+
## Prompt Format
|
38 |
+
The prompt format is as follows:
|
39 |
+
|
40 |
+
```javascript
|
41 |
+
{template keywords} + [company name] + [concept & country] + [industry] + {template keywords}
|
42 |
+
```
|
43 |
+
|
44 |
+
## For example:
|
45 |
+
|
46 |
+
Colored logo
|
47 |
+
```text
|
48 |
+
Simple elegant logo for Google, G circle United states, technology, successful vibe, minimalist, thought-provoking, abstract, recognizable
|
49 |
+
```
|
50 |
+
monochrome logo (include "black and white")
|
51 |
+
```text
|
52 |
+
Simple elegant logo for Google, G circle United states, technology, successful vibe, minimalist, thought-provoking, abstract, recognizable, black and white
|
53 |
+
```
|
54 |
+
|
55 |
+
Here are some examples of prompts:
|
56 |
+
|
57 |
+
- Simple elegant logo for Digital Art, **D A square**, **education**, successful vibe, minimalist, thought-provoking, abstract, recognizable
|
58 |
+
- Simple elegant logo for 3M Technology, **3 M square United states**, **technology and product**, successful vibe, minimalist, thought-provoking, abstract, recognizable
|
59 |
+
- Simple elegant logo for 42Studio, **lines drop fire flame water**, **design studio**, successful vibe, minimalist, thought provoking, abstract, recognizable, relatable, sharp, vector art, even edges
|
60 |
+
|
61 |
+
|
62 |
+
## The [concept & country] section can include words such as:
|
63 |
+
- lines
|
64 |
+
- circles
|
65 |
+
- triangles
|
66 |
+
- dot
|
67 |
+
- crosses
|
68 |
+
- waves
|
69 |
+
- square
|
70 |
+
- letters (A-Z)
|
71 |
+
- 3D
|
72 |
+
- Angled
|
73 |
+
- Arrows
|
74 |
+
- cube
|
75 |
+
- Diamond
|
76 |
+
- Hexagon
|
77 |
+
- Loops
|
78 |
+
- outline
|
79 |
+
- ovals
|
80 |
+
- rectangle
|
81 |
+
- reflection
|
82 |
+
- rings
|
83 |
+
- round
|
84 |
+
- semicircle
|
85 |
+
- spiral
|
86 |
+
- woven
|
87 |
+
- stars
|
88 |
+
|
89 |
+
I invest in cloud GPU rentals to train my models. If you value what I do, consider buying me a coffee ☕️. Your support means a lot! https://www.buymeacoffee.com/iamkaikai666
|